3.1
Motivation
In tradition compressible flow classes there is very little discussion about speed
of sound outside the ideal gas. This author think that approach has many shortcomings. In recent consultation an engineer1 A design of industrial system that
contains converting diverging nozzle with filter to remove small particles from air.
The engineer was well aware about the calculation of the nozzle. Thus, engineer
was able to predict that was a chocking point. Yet, the engineer was not ware what
is the effect of particles on the speed of sound. Hence, the actual flow rate was
only half of his prediction. As it will shown in this chapter, the particles reduces the
speed of sound by almost half. With the new knowledge of the consultation the
calculations were within the range of acceptable results.
The above situation is not unique in the industry. It should be expected
that engineers know how to manged this situation of non pure substances (like
clean air). The fact that the engineer know about the chocking is great but it is
enough for today sophisticated industry2 . In this chapter an introductory discussion
about different situations that can appear the industry in regards to speed of sound.
3.2
Introduction
1 Aerospace
25
26
The people had recognized for several hundred years that sound is
sound wave
dU
velocity=dU
a variation of pressure. The ears
c
sense the variations by frequency
P+dP
P
+d
and magnitude to transfer to the
brain which translates to voice.
Thus, it raises the question: what is
Fig. 3.1: A very slow moving piston in a still gas
the speed of the small disturbance
travel in a quiet medium. This velocity is referred to as the speed of sound.
To answer this question consider a piston moving from the right to the
left at a relatively small velocity (see Figure 3.1). The information that the piston is
moving passes thorough a single pressure pulse. It is assumed that if the velocity
of the piston is infinitesimally small, the pules will be infinitesimally small. Thus,
the pressure and density can be assumed to be continuous.
In the control volume it is convenient to look
Control volume around
at a control volume which
the sound wave
c-dU
c
is attached to a pressure
P+dP
pulse. Applying the mass
P
+d
balance yields
(3.1)
or when the higher term
is neglected yields
(3.2)
From the energy equation (Bernoullis equation), assuming isentropic flow and
neglecting the gravity results
!"
#
%$
'&
(3.3)
(
Substituting the expression for
+*
%$
)&
(3.4)
-,
$
%$
(3.5)
27
An expression is needed to represent the right hand side of equation (3.5). For
ideal gas $ is a function of two independent variables. Here, it is considered that
$ $
where is the entropy. The full differential of the pressure can be
expressed as follows:
%$
$
$
(3.6)
In the derivations for the speed of sound it was assumed that the flow is isentropic,
therefore it can be written
$
$
(3.7)
Note that the equation (3.5) can be obtained by utilizing the momentum
equation instead of the energy equation.
Example 3.1:
Demonstrate that equation (3.5) can be derived from the momentum equation.
S OLUTION
The momentum equation written for the control volume shown in Figure (3.2) is
$
%$+
$
"
+
%
(3.8)
%$
! #
"$ #
$
&
" #
&
%
'
(3.9)
(3.10)
3.3
The speed of sound can be obtained easily for the equation of state for an ideal
gas (also perfect gas as a sub set) because of a simple mathematical expression.
The pressure for ideal gas can be expressed as a simple function of density, , and
a function molecular structure or ratio of specific heats, & namely
'('*)
(+-, ) +/.
10
(3.11)
28
and hence
%$
.
&
('*)
(+-, ) +/.
10
&
Remember that $
written as
(3.12)
+-, ) + .
$
.
&
'*)
&
(3.13)
Example 3.2:
#
Calculate the speed of sound in water vapor at &
steam table (b) assuming ideal gas.
S OLUTION
The solution can be estimated by using the data from steam table3
$ "!$# &%(')#*%
#
0/ 4 1
At & and & : s = 6.9563 + . -, 0/21
= 6.61376 + 35
At 67 and & : s = 7.0100 + . -, 0/ 1
= 6.46956 + 3 0/ 4 1
0/
At 67 and && : s = 6.8226 + . -, 0/21
= 7.13216 + 3 4 1
(3.14)
>=
&
9 7 :*7 9? + @ A 1
(3.15)
for ideal gas assumption (data taken from Van Wylen and Sontag, Classical Thermodynamics, table A 8.)
B
>C 6 9 # : EDF 6 &
&
#
:*
:8:269 + @ A 1
#
Note that a better approximation can be done with a steam table, and it
will be part of the future program (pottoGDC).
3 This data is taken form Van Wylen and Sontag Fundamentals of Classical Thermodynamics 2nd
edition
29
Example 3.3:
The temperature in the atmosphere can be assumed to be a linear function of the
height for some distances. What is the time it take for sound to travel from point
A to point B under this assumption.?
S OLUTION
The temperature is denoted at A as and temperature in B is . The distance
between A and B is denoted as .
Where the distance is the variable distance. It should be noted that velocity is
provided as a function of the distance and not the time (another reverse problem).
For an infinitesimal time
+ is equal to
+
C
&
=
&
#
&
4
6
,
(3.16)
+
B
(3.17)
&
"! %
+
3.4
*
,
4
6
%
(3.18)
.
is it reasonable to put a
discussion here about atmosphere and other affects on
the air?
The ideal gas model can be improved by introducing the compressibility factor. The
compressibility factor represent the deviation from the ideal gas.
Thus, a real gas equation can be expressed in many cases as
$
(3.19)
30
The speed of sound of any gas is (3.7). To obtain the expression for gas that obey
the law expressed by (3.19) some mathematical expression are needed. Recalling
from thermodynamics, the Gibbs function (3.20) is used to obtain
$
(3.20)
*
(3.21)
,
,
(3.22)
31
(3.23)
$ and thus
6
*
*
*
(3.24)
$
, $
,
Substitute the equation
(3.23) results
(3.24) into equation
$
* *
"
$
$
%$
,
*
$
$
6
(3.25)
*
,
$
*
$
%$
,
(3.26)
,
6
(3.27)
Letting
)& for isentropic process results in
$
$
(3.28)
B
in terms of
and
as following
4 See
$#
,
&%
(3.29)
32
Equating the right hand side of equations (3.28) and (3.29) results in
*
,
%$
$
%$
$
#
(3.30)
(3.31)
%
If the terms in the square parentheses are constant in the range under the interest
in this study equation (3.31) can be integrated. For short hand writing convenience,
)
is defined as
0
)
,
and when
$
$
(3.32)
Equation (3.33) the similar to equation (3.11). What is different in these derivation
the relationship between coefficient ) to & was established. The relationship (3.33)
isnt new, and infact any thermodynamics book show this relationship. But with the
definition of ) in equation (3.32) provide a tool to estimate ) In the same manner
as the ideal gas speed of sound the speed of sound for real gas can be obtained.
%$
(3.34)
689 D : 689 D
6
7:
Example 3.4:
,
Calculate the speed of sound of air at &
and atmospheric pressure #
.
&
)
&
The specific heat for air is &
,
, and &
.
Make the calculation based on the ideal gas model and compare these
#
& .
calculation to real gas model (compressibility factor). Assume that
9
S OLUTION
According to the ideal gas model the speed of sound should be
86 9 D & : # 27 : && D 7 9 6 A
@
&
D
For the real gas first the coefficient
69 has
689 D &
& 9
A # 7 :
&& DF 9;: @ A
@
&
*
33
only
3.5
9
4,000 meters, the pressure is about .
. The fractional volume change
is only about 1.8% even under this pressure nevertheless it is a change.
The compressibility of the substance is the reciprocal of the bulk modulus. The amount of compression of almost all liquids is seen to be very small as
given in Table (3.5). The mathematical definition of bulk modulus as following
D 6
9 6
@
$
(3.35)
For example for water
A
)
A
A
A
,1 +
*+ ,
'
'
# #
!
&
*+
(3.36)
6D@
This agrees well with the measured speed of sound in water, 1482 m/s
at 20 C. Many researchers have looked at this velocity, and for purposes of comparison it is given in Table (3.5)
The effect of impurity and temperature is relatively large, as can be observed from the equation (3.37). For example, with an increase of 34 degrees
from 0 C there is an increase in the velocity from about 1430 m/sec to about 1546
[m/sec]. According to Wilson5 , the speed of sound in sea water depends on temperature, salinity, and hydrostatic pressure.
Wilsons empirical formula appears as follows:
$+ '
(3.37)
5 J. Acoust. Soc. Amer., 1960, vol.32, N 10, p. 1357. Wilsons formula is accepted by the National
Oceanographic Data Center (NODC) USA for computer processing of hydrological information.
34
Remark
Fresh Water (20 C)
reference
Cutnell, John D. & Kenneth W.
Johnson. Physics. New York:
Wiley, 1997: 468.
The World Book Encyclopedia.
Chicago: World Book, 1999. 601
Handbook of Chemistry and
Physics. Ohio: Chemical Rubber
Co., 1967-1968: E37
Value [m/sec]
1492
1496
1494
6 D8D 9 6 D
where
is about clean/pure water, is a function tempera
ture, and
is a function salinity, is a function pressure, and is a correction
factor between coupling of the different parameters.
material
Glycerol
Sea water
Mercury
Kerosene
Methyl alcohol
Carbon tetrachloride
reference
25 C
Value [m/sec]
1904
1533
1450
1324
1143
926
Table 3.2: Liquids speed of sound, after Aldred, John, Manual of Sound Recording, London:
Fountain Press, 1972
3.6
The situation with solids is considerably more complicated, with different speeds
in different directions, in different kinds of geometries, and differences between
transverse and longitudinal waves. Nevertheless, the speed of sound in solids is
larger than in liquids and definitely larger than in gases.
Youngs Modulus for a representative value for the bulk modulus for steel
is 160 & N / .
Speed of sound in solid of steel, using a general tabulated value for the
bulk modulus, gives a sound speed for structural steel of
35
reference
longitudinal wave
transverse shear
longitudinal wave (extensional
wave)
Value [m/sec]
12000
5640
5790
3100
5000
Iron
Aluminum
Brass
Copper
Gold
Lucite
Lead
Rubber
5130
5100
4700
3560
3240
2680
1322
1600
Table 3.3: Solids speed of sound, after Aldred, John, Manual of Sound Recording, London:Fountain Press, 1972
D B6 # @
(3.38)
Compared to one tabulated value the example values for stainless steel
lays between the speed for longitudinal and transverse waves.
3.7
The gas flow in many industrial situations contains other particles. In actuality,
there could be more than one speed of sound for two phase flow. Indeed there is
double chocking phenomenon in two phase flow. However, for homogeneous and
under certain condition a single velocity can be considered. There can be several
models that approached this problem. For simplicity, it assumed that two materials
are homogeneously mixed. Topic for none homogeneous mixing are beyond the
scope of this book. It further assumed that no heat and mass transfer occurs
between the particles. In that case, three extreme cases suggest themselves:
the flow is mostly gas with drops of the other phase (liquid or solid), about equal
parts of gas and the liquid phase, and liquid with some bubbles. The first case is
analyzed.
The equation of state for the gas can be written as
'
'
'
'
(3.39)
36
33
6
'
(3.40)
where
' 6 @
3 3
(3.41)
6 @
(3.42)
'
)'*)
0
+-, ) +
(3.43)
Assuming that partial pressure of the particles is constant and applying the second
law for the mixture yields
0! &%
&
/'
$
$
%$
$
(3.44)
$
)'*)
where
Recalling that
6
+-, ) +
(3.45)
(3.46)
-
@
@
(3.47)
33
correction factors for the specific heat is not linear.
37
33
C 3
(3.48)
It can be noticed that and are smaller than similar variables in a pure
gas. Hence, this analysis results in lower speed of sound compared to pure gas.
Generally, the velocity of mixtures with large gas component is smaller of of the
pure gas. For example, the velocity of sound in slightly wed steam can be about
one third of the pure steam speed of sound.
Meta
or a mixture of two phases, speed of sound can be expressed as
where
$
(3.49)
is defined as
Meta End
$
$
(3.50)
38